• ¹ University Politehnica of Bucharest

Instant messenger (chat) is one of the means of Computer Supported Collaborative Learning, which may foster students' inter-animation towards joint knowledge construction. However, the analysis and grading of such chats is an extremely time-consuming activity, therefore (semi-)automatic analytics tools are needed. Discourse threads in a chat may become very complex and intertwined, becoming like voices in a polyphonic musical piece. Starting from the polyphonic model, several learning analytics systems were implemented, in which voices and their inter-animation visualization have a central role. This paper tries to analyze the classes of implemented visualizations, from an interdisciplinary perspective.